The Science of Language

In this Sunday's "On Language" column in the New York Times Magazine (already available online here), I take a look back at the legacy of the column's founder, William Safire. As I write there, "Safire's acute awareness of the limits of his own expertise was often lost on fans and critics alike." Indeed, the "language maven" title that he liked to use was intended to be self-deprecating. (Some might say "self-depreciating," but let's not open that can of worms.)

Part of that self-awareness was a willingness to acknowledge his errors in judgment. In that spirit, I follow up the "On Language" tribute with my latest Word Routes column on the Visual Thesaurus, taking a look at one of Safire's early miscues: declaring, in 1979, that could care less was a "vogue phrase" on its way to extinction. Thirty years later, the verdict is: not so much. Fortunately, Safire didn't often confuse his language mavenry with futurology.

The Reverend Angela Tilby ended her scandalously unresearched little "Thought for the Day" talk of 1 October 2009 (part of which I have already discussed in this recent post) by suggesting that during the British political party conference season (i.e., right about now) we should try taking a blue pencil and editing out all the adjectives from the political speeches so that we could "see what is really being said about people, places, things, deeds and actions". She holds to the ancient nonsense about how nouns tell us the people, places, and things while verbs give us the deeds and actions but adjectives give us nothing but qualifications and hot air and spin — they contribute no content. And she is clearly implying that she (cynically) expects political speeches to be full of adjectives. But as before, she hasn't done any checking at all, she has just spouted her conjectures straight into the microphone. So let's try a second breakfast experiment, shall we?

I examined the first few paragraphs of the transcript of Prime Minister Gordon Brown's speech to the 2009 Labour Party Conference the other day (curiously, he seems to have begun with the coordinator and). Here is the first part of the text, with the adjectives underlined (and again, I am counting them very conservatively, ignoring many items that traditional grammars include under the adjective heading):

And so today, in the midst of events that are transforming our world, we meet united and determined to fight for the future.

Our country confronts the biggest choice for a generation. It's a choice between two parties, yes. But more importantly a choice between two directions for our country.

In the last 18 months we have had to confront the biggest economic choices the world has faced since the 1930s.

It was only a year ago that the world was looking over a precipice and Britain was in danger. I knew that unless I acted decisively and immediately, the recession could descend into a great depression with millions of people's jobs and homes and savings at risk.

And times of great challenge mean choices of great consequence, so let me share with you a little about the choices we are making.

The first choice was this: whether markets left to themselves could sort out the crisis; or whether governments had to act. Our choice was clear; we nationalised Northern Rock and took shares in British banks, and as a result not one British saver has lost a single penny. That was the change we chose. The change that benefits the hard working majority, not the privileged few.

And we faced a second big choice — between letting the recession run its course, or stimulating the economy back to growth. And we made our choice; help for small businesses, targeted tax cuts for millions and advancing our investment in roads, rail and education. That was the change we chose - change that benefits the hard working majority and not just a privileged few.

And then we had a third choice, between accepting unemployment as a price worth paying, or saving jobs. And we in Britain made our choice, it's meant half a million jobs saved. And so, Conference, even in today's recession there are 29 million people in work. 2 million more men and women providing for their families than in 1997.

That's 23 adjectives in 332 words, or 6.9 percent. Decisively less than the scientific paper analyzed earlier, and roughly the frequency one would expect from any ordinary text.

What the Rev. Tilby says is that we should try deleting all the adjectives, which is really absurd (though in fact it is exactly what Alistair Cooke seems to have thought, delusionally, that he used to do to all his radio scripts).

The self-appointed writing gurus who preach in these extreme terms against adjectival modification seem to forget that sometimes adjectives are there because they are crucial not only to the sense but to the structure. Delete the adjectives in this sentence of Brown's and you get a result that doesn't even seem grammatical, and certainly doesn't have anything like the truth conditions of the original. These two are not synonymous:

In the last 18 months we have had to confront the biggest economic choices the world has faced since the 1930s.
In the 18 months we have had to confront the economic choices the world has faced since the 1930s.

Another example, where the result is not even grammatical:

[T]he challenge of change demands nothing less than a new model for our economy, a new model for a more responsible society and a new model for a more accountable politics.
*The challenge of change demands nothing less than a model for our economy, a model for a more society and a model for a more politics.

To make any sense of the claim that such prose could be improved by removing adjectives one would have to propose completely removing all traces of the adjective phrases to which they belong. That would give us the following:

The challenge of change demands nothing less than a model for our economy, a model for a society and a model for a politics.

Why is the Rev. Tilby suggesting that we would understand his proposals better if he couldn't draw the distinction between models and new models, between societies and responsible societies, between politics and accountable politics?

Not that the Prime Minister would have been totally unable to convey his drift, of course. He could in principle have rephrased using only abstract nouns, thus completely avoiding the anti-adjective critique:

[T]he challenge of change demands nothing less than a model for our economy that has novelty, a model (with novelty) for a society that has responsibility to an extent exceeding the responsibility of society as it now exists and a model (with novelty) for a politics with a degree of accountability that exceeds the degree of accountability that politics has today.

Is the Rev. Tilby expecting us to believe that this is an improvement, bringing greater clarity? Has she completely lost her wits? Or did she simply not give any thought to what she was saying?

The notion that you can better see what is being said when the adjectives are removed is simply (yes, I do have to use an adjective here) asinine. Gordon Brown says at one point:

[T]hese are my values — the values I grew up with in an ordinary family in an ordinary town. Like most families on middle and modest incomes we believed in making the most of our talents.

Deleting the adjectives from it yields this:

These are my values — the values I grew up with in a family in a town. Like most families on incomes we believed in making the most of our talents.

What is the point of this ridiculous pretense that it would be a better political world if Brown were blocked from distinguishing ordinary families from unusually affluent ones, not allowed to draw the distinction between having an income and having a median-level income?

Here's why I bothered to write anything at all about a pathetic little 500-word radio sermon: I am so sick of seeing stupid writing advice handed out by pusillanimous pseudo-experts on language — dim-witted vicars like Angela Tilby, pontificating authoritarians like E. B. White in the chapter he added to The Elements of Style, and all the English teachers who have (while hypocritically making free to constantly using adjectives in their own writing) poisoned the reputation of adjectives down the centuries (see the first chapter of Ben Yagoda's delightful little book on the parts of speech, When You Catch an Adjective, Kill It).

These people are wasting educational time and effort, and helping to drive students into a state that I have written about before, characterized by "vague unease instead of a sense of mastery," and feeling "less sure of themselves, yet no better informed," so that their writing ability is "probably being harmed rather than enhanced" — in short, a state of nervous cluelessness about language.

Repeating the falsehood that adjectives are bad in general makes people less able to see what is wrong when they really are over-used. For a remark about the real lesson of Dan Brown's over-use of adjectives, with a diagnosis of what is wrong, see my piece "He doesn't trust us" on the New York Magazine site. There's a real point to be made, I think; but it's not about the adjective category per se.

Adjectives are neither good nor bad. The dumb usage pundits who recommend eschewing them totally are handing out advice that is at best exactly what Angela Tilby wrongly claims adjectives are (vapid, empty, and superfluous), and at worst clearly mistaken.

Josh Fruhlinger sends along today's entry in the "crash blossom" sweepstakes, a headline from the BBC News website:

SNP signals debate legal threat

Crash blossoms (as we've discussed here and here) are infelicitously worded headlines that cause confusion due to a garden-path effect. Here we begin with SNP, which British readers at least will recognize as the abbreviation for the Scottish National Party. Then comes signals, which can be a plural noun or a singular present verb; following a noun, most readers would expect it to work as a verb. The third word, debate, can be a singular noun or a plural verb, and if you've parsed the first two words as Noun + Verb, then you'll be inclined to take debate as the direct object of the verb. So far, so good. But then comes legal threat. What to do now?

Well, you could go back to the beginning of the headline for a reparsing, now construing signals as a plural noun modified by SNP. That would allow you to continue on with debates as a plural verb and legal threat as the object of the verb. But what in the world are SNP signals and why are they debating a legal threat?

Turns out the first path was moving in the right direction. Signals is indeed the verb here, and the object of the verb is debate legal threat — one of those wonderfully opaque compound nouns that British headlines are prey to. You see, debate legal threat refers to threatened legal action that could be taken if the SNP isn't permitted to take part in televised debates before the next UK election. And the SNP is now signaling that it may follow through on this threat.

We've had fun with such outrageous compounding in previous posts (Geoff Pullum in "Noun noun noun noun noun verb," "Canoe wives and unnatural semantic relations," and "Dentist fear girl," and Mark Liberman in "UK death crash fetish?"). This one's a bit different in that the second element of the compound noun, legal threat, is a noun phrase consisting of an adjective modifying a noun. That makes debate legal threat unusually hard to parse.

Noun-Adjective-Noun compounds are possible in English, of course — think of such constructions as Minnesota Supreme Court, Obama White House, Guardian front page, or Microsoft legal team. In those cases, however, the Adjective-Noun component is a set phrase (Supreme Court, White House, front page, legal team), which makes the addition of a premodifying noun unproblematic. But legal threat is not such a set phrase, and debate is not an immediately obvious choice for an attributive noun ready for grafting (certainly not compared to the proper nouns in my examples: Minnesota, Obama, Guardian, Microsoft). So these factors, plus the ambiguous syntactic role of the preceding word, signals, conspire to make this crash blossom particularly crashy.

"Thought for the Day" is a four-minute reflective sermon delivered each morning on BBC Radio 4 at about ten to eight by some representative of one of the country's many religious faiths. On the first day of October the speaker was the Reverend Angela Tilby, Vicar of St Bene't's in Cambridge, England. (Bene't is an archaic shortened form of Benedict.) Developing a familiar theme from prescriptivist literature, she preached against adjectives. It was perhaps the most pathetic little piece of inspirational prattle I have ever heard from the BBC (read the whole misbegotten text here).

"Adjectives advertise," claims the Rev. Tilby, and "brighten up the prose of officialdom", but she was always "encouraged to be a bit suspicious" of them when she was a girl: "Rules of syntax kept them firmly in their place" (as if the rules of syntax left everything else to do what it wanted!). This was good, she seems to think, because "For all their flamboyance they don't really tell you much." Adjectives "float free of concrete reality" like balloons, and are guilty of "not delivering anything except, perhaps, hot air." Which aptly describes her babbling thus far. But now, inflated with overconfidence, she risks some factual statements. And steps from the insubstantial froth of metaphor into the stodgy bullshit of unchecked empirical claims about language use.

I shall deal with only one such claim in this post. Another will be dealt with later.

Because adjectives are so airy-fairy, the Rev. Tilby holds, "you don't find many adjectives in scientific prose and when you do they are precise and exact." I'm sure that Language Log readers will realize instantly that it is time for what Mark Liberman calls a breakfast experiment.

Keep in mind, as I undertake the experiment, that in most kinds of English prose about 6% of the words are adjectives (see Douglas Biber et al., Longman Grammar of Spoken and Written English, London: Longman, 2002, p. 506). In academic prose it's a little higher, around 8%.

I turned to the home page of what is arguably the most important general science journal in the world, Nature, picked the second article title from the top of the page ("Cheater resistance is not futile", by Anupama Khare, Lorenzo A. Santorelli, Joan E. Strassmann, David C. Queller, Adam Kuspa, and Gad Shaulsky, doi:10.1038/nature08472; it just looked somewhat more interesting to me than the first one), and did just a little bit of counting.

You'll notice that the last word of the title (futile) happens to be an adjective, so that's 20% in the title. The first word of the opening sentence of the abstract (cooperative) is also an adjective, and so is the second, and so is the 5th (that's over 15% so far). Here is the whole of the abstract, with the adjectives underlined (I've been very conservative, not counting many items that traditional grammars classify as adjectives: articles, demonstratives, numerals, other determinatives, genitive pronouns, or nouns functioning as attributive modifiers):

Cooperative social systems are susceptible to cheating by individuals that reap the benefits of cooperation without incurring the costs. There are various theoretical mechanisms for the repression of cheating and many have been tested experimentally. One possibility that has not been tested rigorously is the evolution of mutations that confer resistance to cheating. Here we show that the presence of a cheater in a population of randomly mutated social amoebae can select for cheater-resistance. Furthermore, we show that this cheater-resistance can be a noble strategy because the resister strain does not necessarily exploit other strains. Thus, the evolution of resisters may be instrumental in preserving cooperative behaviour in the face of cheating.

That's over 9% adjectives. A bad sample? I took the first paragraph of the text and did the same:

Dictyostelium cells propagate as unicellular amoebae in the soil. Upon starvation, they aggregate into multicellular structures and differentiate into viable spores and dead stalk cells. Stalk-cell differentiation supports spore maturation and dispersal, but this altruistic behaviour can be exploited by cheaters that make more than their fair share of spores in chimaeric fruiting bodies. The genetic potential for cheating is high and cheaters abound in nature, but cheating behaviour can be restrained by various mechanisms, such as intrinsic lower fitness of the cheater, pleiotropy of the cheater gene, high genetic relatedness in natural populations, and kin discrimination.

That's 16 adjectives in 97 words of that paragraph, or over 16%. In total, the title and abstract and opening paragraph of the first scientific paper that I picked — genuinely a random choice — are nearly 13% composed of adjectives, well over double the frequency that you find in most prose.

Now, I could check a few hundred more words, of course. But wait: why me? Why am I doing the work for her? What am I, an unpaid assistant curate of St. Bene't's? Did the Rev. Tilby do even as much elementary checking as I have done so far — glancing at a couple of hundred words in a random paper — before spouting her ridiculous remark? Of course not. Her method is a time-honored one in amateur writing on language: she just makes stuff up. On the basis of nothing but prejudice about science, she invented her data and went straight to the microphone with it.

Her suggestion that in science the adjectives are "precise" is further evidence of uninformed stereotyping. There's nothing precise about the meanings of words like cooperative, social, viable, altruistic, fair, lower, high, natural… These are vague terms, in the classic technical sense: in any situation there will be clear cases for their application, but also a border area where the appropriacy of applying them is in doubt.

There is of course nothing wrong with vague terms, with denotations partly set through common sense and reference to context; we use them literally every minute that we speak or write. Their logic and semantics can be studied with ruthless precision (see, for an example of the technical literature, Stewart Shapiro's lucid and masterful Vagueness in Context). Science is replete with them, and has to be. (Think of global warming, for heaven's sake: there's a truly vague concept. How warm? How global? Yet it's an important one, and serious science is being done every day to flesh it out and give it clearer content.)

It is merely one more sign of the the Rev. Tilby's contempt for truth, and cluelessness about science, that she thinks science is all precision. Scientists live their lives floating in a probabilistic soup of uncertainty and unclarity, murky associations and ill-defined tendencies, statistical degrees and extents.

Tilby sees herself as a minister of religion and thus a professional talker; and she therefore assumes (the crucial fallacy) that she is an expert on language; so she doesn't need to check a thing. As a vicar, she thinks she can go into the Radio 4 studio and simply invent her facts.

It is not that she lied; it is worse than that. Tilby didn't know what the facts about adjectives in scientific prose were, and state untruths about them to mislead her audience: she simply didn't care whether she was uttering untruths or not. It wasn't lies; it was bullshit, in the sense defined by Harry Frankfurt. And as Frankfurt notes, the purveyor of bullshit is worse than a liar, in virtue of caring less about truth. The liar at least keeps track of what's true and recognizes its special status. (That is precisely why it is a tangled web we weave when first we practice to deceive: the committed liar has to attempt to remain consistent.)

A recent xkcd:

In the case of power, the original ordinary-language meaning is still dominant for most people, especially in a frame like "With great __ comes great __". But it's easy to forget how recently words (and concepts) like speed, distance, and duration took on their current "literal" meanings as aspects of ordinary-language physics rather than as terms referring to prosperity, dissension, endurance, and so on.

The physicists' sense of power as "work per unit time" seems to date from the early 19th century, and the specifically electrical sense, featured in this strip's caption, is somewhat later. But it was only a century or two earlier that today's meanings for words like distance came into general use, replacing earlier meanings that (like power) had more to do with personal struggle than with physical interaction.

When phrases are coordinated, readers infer that the the juxtaposed elements are in some way parallel. Careless coordination produces unwanted inferences. Today's Daily Beast serves up an object lesson:

Stunned colleagues Friday described veteran CBS News producer Joe Halderman—who was arrested outside the network’s West 57th Street offices Thursday in the alleged scheme to blackmail David Letterman—as a rogue and a womanizer, a lover of literature, a “smart frat boy,” a swashbuckling journalist, and an occasional barroom brawler who distinguished himself in dangerous war zones and occasionally displayed a certain reckless streak.

Fucking literature lovers.

Today's Zits:

Rick Rubenstein wrote:

Is the usage "I can't speak to the Iranian situation" as opposed to "I can't speak [about/regarding] the Iranian situation" relatively recent (or at least recently accelerating), as I perceive it to be? I feel as though I first noticed it about a decade ago, and found it very strange. I'm now almost accustomed to it.

There's no question that "speak to (a topic)" is quite a bit more recent than "speak of (a topic)", and somewhat more recent than "speak about (a topic)". But Rick is probably not old enough to have noticed the difference.

In the OED's entry for speak, the sub-entry II.11.a. Speak of, which is glossed "To mention, or discourse upon, in speech or writing", is cited from about 825:

c825 Vesp. Psalter cxviii. 46 [Ic] sprec of cyðnissum ðinum in ᵹesihðe cyninga. c950 Lindisf. Gosp. Luke ix. 11 [He] spræcc him of ric godes. c1175 Lamb. Hom. 73 Of þe halie fulht spec ure drihten on oðer stude. c1200 ORMIN 6784 Goddspellboc ne spekeþ þ nohht Off all þatt oþerr genge. c1340 HAMPOLE Pr. Consc. 2683 Here es þe thred parte of þis buke spedde Þat spekes of þe dede. 1422 Secreta Secret., Priv. Priv. 203 Of this Spekyth the boke of Iudyth. 1530 PALSGR. 727/2, I go nowe beyondsee, but if God send me lyfe you shall here speke of me. 1603 PARSONS Three Convers. Eng. II. viii. 481, I shall haue occasion to speake againe of these heretiks in the next chapter. 1730 A. GORDON Maffei's Amphith. 58 The Theatre..is spoke of by Martial. 1818 SCOTT Br. Lamm. xviii, ‘And speaking of red-game,’ said the young scape-grace, interrupting his father. 1884 tr. Lotze's Metaph. 43 A common-place with every philosophy which spoke of Things at all.

In contrast, the sub-entry II.14.e. Speak to, glossed "To treat of or deal with, to discuss or comment on, (a subject) in speech or writing", is only cited from 1610, almost eight centuries later:

1610 J. DOVE Advt. Seminaries 42, I desire them therefore..to speake to these foure points. 1637 HEYLIN Answ. Burton 78, For your charges,..I meane to take them..in order, and speake as briefely to them, as you would desire. 1662 STILLINGFL. Orig. Sacræ II. vi. §4 Though it be a subject little spoken to either by Jewish or Christian Writers. 1706 STANHOPE Paraphr. III. 555 Part of this Scripture hath already been spoken to. 1724 SWIFT Drapier's Lett. Wks. 1755 V. II. 110 A lawyer, who speaks to a cause, when the matter hath been almost exhausted by those who spoke before. 1778 EARL MALMESBURY Diaries & Corr. I. 166 Unprepared as he was for such a proposition, he could not, he said, off-hand, speak to it accurately. 1869 Daily News 28 Apr., The report..was spoken to by the Most Rev. Chairman..and the Bishop of Derry. 1880 Ibid. 19 Mar. 2/3, I wish to call your attention..to..that allegation, and I shall endeavour to speak to it.

As for Speak about (sub-entry II.8), it's cited back to 1300 or so, validating Rick's sense of its antiquity:

a1300 Cursor M. 24795 For to spek abute sum pais. 1605 SHAKES. Macb. I. iii. 83 Were such things here, as we doe speake about? 1671 H. M. tr. Erasm. Colloq. 263 He falls on speaking about the success of their business. 1737- [see 14b]. 1843 J. H. NEWMAN Lett. (1891) II. 430 Sermons which speak more confidently about our position than I inwardly feel.

Rick also asks, "I'm also curious which side of the Atlantic this usage may have sprouted from." It seems clear from the OED entry that all the early action was on the British side of the Atlantic.

More seriously, it's quite possible that there's been a recent to-ward change in the balance of usage among the prepositions used with speak to express topic (which include at least of, about, regarding, on, upon, and to, of course with somewhat different shades of meaning and structural distributions).

Unfortunately, it's going to be a chore to test this quantitatively. One obvious problem is that there may be various things between the verb and a prepositional phrase expressing topic, e.g. "Mr. Pettijohn spoke at length regarding the Rocky Top Road issue". Another, more serious, problem is that in most instances of "speak to", the object of to is the audience, not the topic ("Palin Speaks to Investors in Hong Kong"). So (lacking an automatic classifier with adequate performance), you'd have to get a suitable random sample of instances of speak over time, and classify each one by hand. This is likely to take more work than will fit into one Breakfast Experiment™, at least with the resources now available to me.

[Note that the specific pattern "speak to the * situation" is apparently not common enough to support a trend analysis. For example, it has apparently only occurred once in the NYT news archive since 1981. So the net would have to cast more broadly in order to spot a trend, I think.]

First, a new twist on a story that our legal desk covered back in February: at the annual Ig Nobel awards ceremony earlier tonight, the Prize for Literature was awarded to the Garda Síochána na hÉireann (i.e. the Irish Police Force) for the 50 or more speeding tickets they've issued in the name "Prawo Jazdy", Polish for "driver's license."

And as if that wasn't enough onomastic excitement, the 2009 Ig Nobel Prize for Veterinary Medicine was awarded for work reported in Bertenshaw, C. and Rowlinson, P., Exploring Stock Managers' Perceptions of the Human-Animal Relationship on Dairy Farms and an Association with Milk Production, Anthrozoos: A Multidisciplinary Journal of The Interactions of People and Animals 22:1, pp. 59-69, 2009. Specifically, Dr. Bertenshaw and Dr. Rowlinson share the prize for their demonstration that (and here I quote from the article's abstract): "On farms where cows were called by name, milk yield was 258 liters higher than on farms where this was not the case (p < 0.001)."

Yet all this groundbreaking research leaves me with more questions than answers. What is the causal direction behind the correlation? And if my cow produced 238 liters too little milk, would I admit to the researchers the names I used for her? And how much milk can an Irish policeman get from a speeding Polish cow?

Mark Liberman has been writing about the persistent misinterpretation of claims about statistical differences between groups with respect to some property — misinterpretations in which these differences (often small ones) are understood as general (and essential) differences between the groups. A little while ago, Mark suggested that reporting the differences by means of generic plurals ("Asians have a more collectivist mentality than Europeans do") promotes misunderstanding and proposed that such generic language be avoided.

In comments, Mark noted that changing the language of popular science reporting is not by itself going to fix an inclination to essentialist thinking (which is all over the place), though it might be a step in the right direction.

Mark looked at groups X and Y with respect to property P, focussing on statistical differences between X and Y. Let's say that members of X tend to have P to a greater degree than members of Y. Then there's a statistical association (perhaps a very weak one) between X and P. And most people have as much trouble understanding statistical association as they do statistical differences. Here, the characteristic error in reasoning is treating the association as invariant: all Xs have P (and no non-Xs do). Again, the error is all over the place, often showing up in objections to claims of statistical association. As in the following case from a September 30 NYT story "Dementia Risk Seen in Players In N.F.L. Study" by Alan Schwarz.

The story begins with a report of an apparent statistical association:

A study commissioned by the National Football League reports that Alzheimer's disease or similar memory-related diseases appear to have been diagnosed in the league's former players vastly more often than in the national population — including a rate of 19 times the normal rate for men ages 30 through 49.

Now, the study has a variety of shortcomings, and the association can't be taken as demonstrated. but for the moment let's take things at face value.

What especially interests me in the story (though I have a long-standing interest in dementia and related conditions) is the reaction of NFL spokesman Greg Aiello, who

said in an e-mail message that the study did not diagnose dementia, that it was subject to shortcomings of telephone surveys and that "there are thousands of retired players who do not have memory problems."

"Memory disorders affect many people who never played football or other sports," Mr. Aiello said.

Aiello's response addresses (and rejects) claims not made in the report: that all retired players have memory problems (all Xs have P), and that no other people have memory problems (no non-Xs have P). That is, Aiello chose to treat the association as invariant, when that was clearly not what the report said. I don't see how the reporter could have averted this misunderstanding by more careful wording.

Such misunderstandings — occasioning objections much like Aiello's — arise even when claims are not couched as statistical associations. Quantified statements like "many Xs have P" will do, eliciting objections like "but I'm an X and don't have P". People reason like this all the time, and that's why there are Critical Thinking courses.

Andrew Gelman at Statistical Modeling, Causal Inference, and Social Science, commenting on my posts about a Dilbert cartoon that skewers "the vacuous way managers talk", asks "What is a 'manager' anyway?"

My only comment here is not on the Bayesian inference but rather on the idea that "managers" are dweeby Dilbert characters who talk using management jargon. I was thinking about it, and I realized that I'm a manager. I manage projects, hire people, etc. But of course I don't usually think of myself as a "manager" since that's considered a bad thing to be.

For another example, Liberman considers a "spokesperson for a manufacturer of sex toys" as a manager. I don't know what this person does, but I wouldn't usually think of a spokesperson as a manager at all.

The LL posts in question were "Moving low-hanging fruit forward at the end of the day", 9/26/2009;  "Memetic dynamics of summative cliches", 9/26/2009; "'At the end of the day' not management-speak", 9/26/2009; "Another nail in the ATEOTD=manager coffin", 9/28/2009; and "Memetic dynamics of low-hanging fruit", 9/30/2009.

And Andrew's comment is very much to the point.

In the original post on Scott Adams' strip, I used scare quotes in referring to "a despised minority, in this case 'managers'"; and in a later post, I referred to "the hypothesis that it's especially likely to be used by 'managers', however we define that much-maligned class".  Those phrases were meant as a note to myself to come back to this curious phenomenon of anti-managerial prejudice — but Andrew beat me to it.

I should start with a confession: like Andrew, I'm a manager. I've been one at least since 1980 or so, when I became a group supervisor at Bell Labs; and when I became a department head there, in 1987,  it was no longer possible to look the other way.   At my first "three-level meeting" — involving department heads and their bosses and their bosses' bosses — Brian Kernighan looked across the conference table to me and said "Welcome to the land of the brain-dead". The fact that I left industry for academia didn't save me: the organizations that I now (at least nominally) manage employ several times as many people as the Linguistics Research Department at Bell Labs did.

I suppose that as long as there have been hierarchies, bosses have sometimes been feared, resented, and disliked.  But I think it's a new phenomenon that in large areas of modern culture, managers are stereotypically regarded as stupid.  Andrew Gelman's employees surely don't think that he's stupid, and I hope that most of mine don't think I'm stupid either. But Brian's little joke reflected a now-widely-shared concern that the role of manager, like the role of parent, inevitably causes a sort of tragic cluelessness, in which you become the object of all of your own earlier upward-facing attitudes.

This is mixed up with a different idea, namely that official pronouncements are likely to be bland and even empty. This might mean that the people who craft them are actually especially crafty, but  the idea that corporate statements tend toward vacuity seems to  reinforce the idea that leaders are empty-headed.

[It's for that reason that I was disposed to accept corporate spokespersons as "managers" in the Dilbertian sense — well, that and the fact that otherwise there would otherwise have been no examples at all in the 400-million-word COCA database of "managers" using the cliche under study.  And the sex-toy company spokesperson whose quote I accepted as a possible example of manager-speak actually was a "senior buyer" — check out the original article here.]

However we decide to define "manager", this group is certainly now the object of a complex of negative stereotypes. When and how did this start?

I don't know, and I welcome suggestions.  These attitudes may be connected to the antique European aristocratic disdain for those who are "in trade", and to the (I think related) modern intellectual disdain for the world of business.  These attitudes seem to have been imported from the intelligentsia into  industry through the medium of engineers and especially programmers, who (at least at lower levels) maintain a very different culture from the "suits" in finance, marketing, product planning, and so on.

The word manager has been around for a long time, with something close to its current meaning. With the gloss "A person who organizes, directs, or plots something; a person who regulates or deploys resources", the OED gives citations from around 1600 onwards:

1598 J. FLORIO Worlde of Wordes, A manager, a handler. 1598 SHAKESPEARE Loves Labours Lost I. ii. 173 Adue Valoure, rust Rapier, be still Drum, for your manager is in loue. 1600Midsummer Night's Dream V. i. 35 Where is our vsuall manager Of mirth? What Reuels are in hand? SHAKESPEARE

In the slightly more specific sense of  "A person who manages (a department of) a business, organization, institution, etc.; a person with an executive or supervisory function within an organization, etc.", the citations start about a century later:

1682 A. WOOD Life 22 Nov., The Duke of York hath gained the point as to the penny post against Docuray the manager of it. 1705 J. ADDISON Remarks Italy 443 The Manager opens his Sluce every Night, and distributes the Water into what Quarters of the Town he pleases. 1740 S. RICHARDSON Pamela II. 341 Said he, I think that little Kentish Purchase wants a Manager.

But none of the early citations in the OED, nor the quotes that I find in LION, seem to reflect the modern Dilbertian managerial stereotype.  That stereotype clearly predates Dilbert — but when did it arise? and where did it come from?

In this context, we have to return to Andrew's question: What is a manager, anyhow?  By now, I suppose that the Dilbert empire employs a certain number of people, whom Scott Adams in some sense manages — does he thereby consider himself a "manager" in the relevant sense?

The fuzzy referential boundaries of the managerial stereotype, it seems to me, are a characteristic of social stereotypes in general. This is related to the "some of my best friends are Xs" excuse, and all the other excuses that shift the range of the prejudice away from apparent counterexamples.

[Update — John Cowan's post at Recycled Knowledge, "Why are PHBs stupid?", offers interesting and sensible answers to the questions under discussion.]

One of the most powerful pieces in the one-man exhibition of Chen Wen Ling now showing at Joy Gallery in Beijing 798 Art Zone is blandly entitled "What You See Might Not Be Real" in English, but the Chinese title is the raw and raunchy "FANG4PI4″ 放屁 ("emit gas, break wind, flatulate, crepitate, i.e., fart [v.]"). Perhaps the artist didn't want to offend the linguistic sensitivities of potential foreign customers, but I must say that I much prefer to translate the title of the piece directly as "Fart," or, with a bit of license, as "Bull Fart" because the atomic cloud depicted by the artist is coming out of the anus of an enormous bovine.

[This view comes from a story at Business Insider (Joe Weisenthal,  "Finally! Madoff Gets What He Deserves", 9/29/2009); other perspectives are available at ML Art Source and TPM.]

What we see in this impressive sculpture is a horned Bernie Madoff pinned against the wall by the rocket-propelled bull. Just from looking at the whole ensemble, it's pretty obvious what Chen is trying to tell us, but by entitling the piece "FANG4PI4," he invokes additional levels of scorn that are inherent in that term when applied to the words of others. Several of the blogs that have shown this piece claim that as slang FANG4PI4 implies "bluff" or "lie." Actually, it is more accurate to say that it means "talk nonsense," the idea being that one is comparing the words coming out of the mouth of one's opponent to a stream of farts.

Now, if one wishes to increase one's contempt for what one's opponent is saying, one may style his / her words as GOU3PI4 狗屁 ("dog fart"), as in this ringing denunciation: NI3 FANG4 GOU3PI4! (lit. "You are emitting dog farts!" = "What you say is nonsense / bullshit!"), although GOU3PI4 ("dog fart[s]!") shouted loudly by itself gets the message across clearly enough. This is an old expression that may be found as early as 1750 in the Qing Dynasty novel Rulin waishi (The Scholars). If you want to emphasize that what your opponent is saying is not only bullshit but is also completely incoherent, you may declare that it is GOU3PI4 BU4TONG1 ("dog fart not pass through").

If your adversary still does not give in to your withering denunciations, you may embellish them as follows (I shall only give a few of the possible varieties):

FANG4 GOU3CHOU4PI4 ("emit stinking dog fart[s]")

FANG4 NI3 MA1 DE 4PI4 ("emit 'your mother's' fart[s]")

FANG4 NI3 MA1 DE GOU3CHOU4PI4 ("emit your mother's stinking dog fart[s]")

However, one must be careful when one gets into the territory of "your mother's" whatever, since such characterizations are considered to be extremely vulgar and, as often as not, fighting words. We all remember the "Grass Mud Horse" phenomenon from earlier this year, and a lot more could be said about this most offensive of imprecations.

As for "bull," that is NIU2 牛, although I did mention in an early January post that "Happy NIU2 Year" was the most popular STM New Year's greeting in China this year, I have not yet found the time to explore the full range of nuances of NIU2 in current usage ("balls, guts, spunk, awesome, formidable," and so forth). Particularly when combined with "B," viz., 牛B (often written as NB), then we begin to combine all of the androgenic qualities of NIU2 with the estrogenous implications of what "mother's" refers to, resulting in an explosive combination. To do full justice to this aspect of NIU2 would require a modest (or perhaps I should say "immodest") treatise, one that I have not yet found the opportunity to compose. Someday.

[Hat tip to Benjamin Zimmer.]

Commenting on a post about Dilbert's take on "the vacuous way managers speak", Garrett Wollman wrote:

I remember, or at least think I do, when "low-hanging fruit" was not yet vacant managerese. Is there any epidemiological data to suggest when this transition occurred?

I'm not convinced that "low-hanging fruit" is accurately described as "vacant managerese" even now. But let's leave that point aside while we consider the epidemiological data on the rise of this cliche among all classes of users, which suggests an index case in the late 1980s, with the main contagion starting in the mid 1990s:

This graph is based on data from the New York Times archive, and is derived from the counts in the following table, which tracks occurrences of "low-hanging fruit" and four other cliches, namely "easy pickings", "shooting fish in a barrel", "easy as pie", and "a piece of cake":

TIME LHF EPick SFB EPie PCake 1981-1985 0 23 4 11 72 1986-1990 2 17 11 6 90 1991-1995 4 33 15 5 67 1996-2000 34 34 21 12 84 2001-2005 74 32 16 7 78 2006-2009 78 14 17 5 41

In order to correct for possible changes in number of words indexed per time-period, I've added up the counts for the four other cliches, and graphed the ratio of "low-hanging fruit" to that sum. Obviously, in this case, a graph of the raw LHF counts would show a similar pattern.

Garrett added:

I also remembered when other geeks recommended Malcolm Gladwell's The Tipping Point to me, which now seems to be exclusively marketed to management types who understand it not.

The "marketing types" here at Penn are Wharton students, whose level of technical understanding is second to few, I think.  (Though I certainly recall from my years in industry that developers viewed "marketing types" as being an especially clueless subspecies of "suits", less actively evil than finance types, but stupider.  I'm sure that the feelings were reciprocated, mutatis mutandis.)

But in any case, since Gladwell's book was published in 2000, well into the spread of the LHF epidemic, I don't think that he deserves either any blame or any credit for the process.  Another clue to his innocence is the fact that the string "low-hanging fruit" (with or without the hyphen) doesn't occur anywhere in the book. (Nor, for  that matter, does "at the end of the day"…)

[Note — the OED's earliest citation is

1990 N.Y. Times 16 Aug. D1/3 We've picked all the *low-hanging fruit when it comes to fuel efficiency.

That's indeed the earliest citation in the NYT's archive. It's not hard to back that up by a bit, e.g. Brenton Schlender, "Some U.S. Makers of Semiconductors Are More Optimistic", WSJ 2/17/1987:

I expect that (figurative) examples can be found from several years earlier — the phrase in its literal meaning, of course, goes back hundreds of years. The earliest literal use I've found is from Adam Bede, 1859:

He could see there was a large basket at the end of the row: Hetty would not be far off, and Adam already felt as if she were looking at him. Yet when he turned the corner she was standing with her back towards him, and stooping to gather the low-hanging fruit. Strange that she had not heard him coming! perhaps it was because she was making the leaves rustle. She started when she became conscious that some one was near—started so violently that she dropped the basin with the currants in it, and then, when she saw it was Adam, she turned from pale to deep red. That blush made his heart beat with a new happiness. Hetty had never blushed at seeing him before.

But I think it would be surprising if there weren't earlier examples as well.]

[Update — Jesse Sheidlower sent in an earlier use as an explicit simile:

1968 P. J. Kavanaugh in _Guardian_ 12 July 6/3 His work is so appealing to me that I feel almost bashful praising it… He is gentle and stoic and simple, his rare images are picked aptly, easily, like low-hanging fruit, and though he appears to move short distances slowly he really moves far and fast.

Jesse's find suggests looking for even earlier active similes and metaphors, perhaps with slightly different wording, and indeed there are many. Thus Wilfred Scawen Blunt, Griselda (1914):

57   For we, in truth, were no wise company,
58   Men strong and joyous, keen of hand and eye,
59   And shrewd for pleasure, but whose subtlest wit
60   Was still to jest at life while using it,
61   And jest at love, as at a fruit low hung
62   To all men's lips, no matter whence it sprung.

Or George MacDonald, To Any Friend (1893):

9   At home, no rich fruit, hanging low,
10      Have I for Love to pull;
11   Only unripe things that must grow
12      Till Autumn's maund be full!

Or Dora Sigerson, My Darling (1893):

5   Oh, Life came over the meadows,
6      And the song of her coming was sweet;
7   The streams leaped joy-mad down the mountains,
8      Flowers bloomed 'neath her dawning feet.
9   The trees bent their branches fruit-laden,
10      So low as her soft hands' hold;
11   And the harvest rose up like an army
12      Of kings in their harness of gold.

Or Edmund Gosse, VIllanelle I (1879):

1   Wouldst thou not be content to die
2      When low-hung fruit is hardly clinging,
3   And golden Autumn passes by?

Or Herman Melville, Clarel (1876):

3919   Estranged, estranged: can friend prove so?
3920   Aloft, aloof, a frigid sign:
3921   How far removed, thou Tree divine,
3922   Whose tender fruit did reach so low—
3923   Love apples of New-Paradise!

OrAaron Hill (died 1750), Sareph and Hamar:

42      Refresh'd by sleep, he rose serene and gay,
43      And walk'd abroad to see the breaking day,
44   With dawning lustre, thro' the boughs, in trembling sallies play.
45      Where-e'er he pass'd, the golden fruit hung low
46      And dancing, wanton, bow'd to court his hand,
47      Proud of the native charms they had to show;

Or  Charles Goodall, To Mr. R. Smith of King 's Colledge in Cambridge (1689):

33   The barren Tree can in the Desarts spread,
34   And threaten Heaven with its luxurious head:
35   Whilst others low, and laden with their Fruit,
36   With bended Branches touch their very root.

Or Henry Reynolds' 1628 translation of Torquato Tasso's Aminta:

Being but a Lad, so young as yet scarse able
To reach the fruit from the low-hanging boughes
Of new growne trees; Inward I grew to bee
With a young mayde, fullest of loue and sweetnesse,
That ere display'd pure gold tresse to the winde;

In fact, you could say that LHF was a poetic cliche long before it came to be a cliche associated with any other group. And I wouldn't be surprised to find the metaphor in Homer, or the Bible.]

On a very interesting and informative blog called "This is Xinjiang", we find the following sign over the entrance to an Ürümchi restaurant:

The blogger, an anonymous "foreign university teacher in China's western frontier", captions the photograph thus:

The diverse number of scripts found on Xinjiang signs — Arabic, Chinese, Mongolian, and Russian — often present a bewildering challenge to Westerners who are used to the Latin alphabet. Here, no one writes English in public places. But Xinjiang's multilingualism can even mislead the locals. For example, this restaurant, located in downtown Urumqi, advertises "Pancakes, Hamburgers, Porridge" in Chinese characters. Unfortunately, the Uyghur "translation" written above that is: ngngoongngkngngnglng.

Who's pulling whose leg?  It is evident that, in this case, the "Uyghur" language written in Arabic script is there for purely decorative purposes, reminiscent of the mirror-reversed Chinese characters in a New Yorker ad sponsored by the Massachusetts Office of Travel and Tourism that I wrote about here, or the Hong Kong strip-club flyer used to decorate the cover of a China-themed issue of MaxPlanckForschung.

My guess is that even someone who doesn't know Arabic script or Uyghur language would be suspicious of the numerous repetitions of just a few letters.

Such signage is supposed to be supervised by the official Working Comittee for the Languages and Scripts of Minority Nationalities (少数民 族语言文字工作委员会 Az sanliq milletler til-yeziq xizmet komiteti).

And this is another way to "read" the Uyghur letters: ng_ng_o_o_ng_ ng_k_ng_ng_ ng_lng.  Try it without choking!